Task-switching paradigms consistently show an age-related increase in mixing cost ie., older adults show a greater increase in RT on repeat trials in mixed-task blocks as compared to single task blocks. We examined changes in mixing cost in four age groups spanning across the adult lifespan using a cued trials switching paradigm. We compared behavioral variables and latent variables derived from evidence accumulation models of speed-accuracy tradeoff against electrophysiological variables measuring preparation for an impending repeat trial and stimulus processing. Increasing age was associated with a higher RT and greater RT variance mixing cost but smaller error mixing cost, suggestive of age-related changes in speed-accuracy tradeoff. Diffusion model parameters indicated a more conservative decision process under mixed-task conditions in older adults. Cue-locked ERPs showed a prolongation of the mixing-positivity across the lifespan and this was associated with greater criterion increase for mixed-repeat trials. Stimulus-locked ERPs showed gradual changes in the mixing effect across the lifespan and this was associated with drift rate reduction for mixed-repeat trials. The results suggest that the early emergence of strategic differences in response criterion may modulate preparatory processes. These do not adversely affect behavioral performance until slowing of the rate of evidence accumulation in older adults makes this strategy ineffective. The analysis strategy used here indicates that diffusion model parameters and ERP measures are complementary approaches that can illuminate the processes that contribute to age-related cognitive changes.
9th Conference of the Australasian Society for Cognitive Science (ASCS 2009). ASCS09: Proceedings of the 9th Conference of the Australasian Society for Cognitive Science (Sydney 30 September - 2 October, 2009) p. 170-177